Machine learning in functional cancer assays

Abstract

The invention provides methods that use machine learning to discover clinical data patterns that are predictive of disease, such as cancer. Clinical data from across a population is provided as input to a machine learning system. The machine learning system discovers associations in data from a plurality of data sources obtained from a population and correlates the associations to cancer status of patients in the population. The methods may further include providing patient data from an individual and predicting, by the machine learning system, a cancer state (e.g., the presence of cancer and a determination of a stage or progression of the cancer, if present) for the individual when the patient data presents one or more of the discovered associations.

What is claimed is: 
     
         1 . A method of evaluating cancer, the method comprising:
 providing, to a computing system, a training data set comprising functional biomarker measurements from a plurality of patient samples each having a known cancer status;   associating the functional biomarker measurements with the cancer statuses;   obtaining a sample from a patient suspected of having cancer;   measuring a functional biomarker of one or more live cells isolated from the sample;   inputting data obtained in the measuring step into the computing system;   correlating the data with the cancer statuses; and   reporting results of the correlating step.   
     
     
         2 . The method of  claim 1 , wherein the measuring step includes measuring a mass or mass accumulation or mass accumulation rate for the one or more live cells. 
     
     
         3 . The method of  claim 2 , wherein the mass or mass accumulation, or mass accumulation rate is measured using a device comprising one or more suspended microchannel resonators. 
     
     
         4 . The method of  claim 1 , wherein the sample comprises a fine needle aspirate or biopsy from a patient suspected of having cancer. 
     
     
         5 . The method of  claim 1 , wherein the computing system comprises a machine learning system selected from the group consisting of a random forest, a support vector machine, a Bayesian classifier, and a neural network. 
     
     
         6 . The method of  claim 1 , wherein the computing system comprises an autonomous machine learning system that associates the functional biomarker measurements with the known cancer statuses in an unsupervised manner. 
     
     
         7 . The method of  claim 6 , wherein the autonomous machine learning system comprises a deep learning neural network that includes an input layer, a plurality of hidden layers, and an output layer. 
     
     
         8 . The method of  claim 6 , wherein the autonomous machine learning system represents the training data set using a plurality of features, wherein each feature comprises a feature vector. 
     
     
         9 . The method of  claim 6 , wherein the autonomous machine learning system comprises a random forest. 
     
     
         10 . The method of  claim 1 , further wherein the training data set includes at least one other source of data associated with the known cancer statuses. 
     
     
         11 . The method of  claim 10 , wherein the at least one other data source includes genomic data. 
     
     
         12 . The method of  claim 1 , further comprising operating a machine learning system to learn relationships among cancer statuses, treatment options, depth of response, known treatment efficacies, and progression free survival. 
     
     
         13 . The method of  claim 12 , further comprising selecting, by the machine learning system, one or more recommended treatments for the patient based, at least in part, on the results of the correlating step and learned relationships. 
     
     
         14 . The method of  claim 12 , wherein one or more of the training data set, cancer statuses, treatment options, depth of response, known treatment efficacies, and progression free survival are obtained from one or more publicly available data repositories.